Gender differences in usage of interrogative utterances in everyday English

(11) S0274: >>running a business basically sh- her position like many of them was bought

S0253: right

S0274: >>you know mini funded her old campaign and stuff like that and she what life experience has she got ?

S0253: yeah

The fact that this interrogative can be reformulated via a declarative “She's got no life experience” illustrates how it, in fact, is a matter of expressive choice and mode of information delivery rather than its request.

In regard to tag questions acting in a rhetorical way, as discussed before, there is trouble with Freed's lack of demonstrative examples of phatic versus rhetorical tag questions. I will thus attempt to elaborate on examples given by Freed by contrasting them. Below is Freed's Example (12) showing a rhetorical tag question:

(12) A: ... Normally I leave 45 minutes before a class. Just to make sure.

B: That long?

A: Yeah. It's too much, isn't it? Then I get here early.

B: I don't get here that early.

Firstly, this interrogative does not elicit an answer, which is said to be characteristic of rhetorical questions [e.g., Kearsley 1976, 359]. This, however, does not exclude the phatic function as such questions are often embedded in the speaker's turn and are also not treated as information requests (see Example 13):

(13) S0254: well no but they 've done

S0253: er

S0254: they do studies do n't they ? and to find like

S0253: okay

S0254: in general people who eat five portions of fruit and veg a day

The key may then lie in how in Example (12), the speaker's point of view is expressed, while in Example (13), background information is brought up, its accuracy, presumably, out of question. The speaker is thus prompting the hearer to agree (not necessarily verbally, but here, it is verbal; “okay”). However, Freed's example of a phatic tag question (“It's interesting, isn't it?”) also seems to express a point of view but it is impossible to say for sure since, sadly, there is no conversational context at all. If the same question were to be placed in the context of Example (13), arguably, it would fit in without disturbing the flow or the interpretation of the utterance. We have no choice but to assume that Freed's example's original context played the determining role in its classification as a phatic question. It can be suggested that the word “interesting” has the capacity to possess more senses than the simple end-of-the-spectrum “too much”. By asking “It's interesting, isn't it?” at the end of the turn, one could have invited the interlocutor to participate, similar to Example (7) in section 2.2.3.3. Or it could have been embedded into the utterance, followed by elaboration on some topic as in Example (13).

We thus assume that a tag question's phatic rather than rhetorical (or vice versa) function is determined by the conversational context and treat them as one or another taking into account the nature of information conveyed/requested - a point of view or background facts. As for how the use of rhetorical function can correspond to gender, not much can be speculated.

2.2.4.3 Humor

Humorous questions are described as expressing information “from an unexpected point of view” [Freed 1994, 631], and unlike didactic or rhetorical questions, they are aimed at sharing amusement. This function is probably the most context-sensitive and, most importantly, intonation-dependent, hence why detecting them in transcribed data is problematic. Questions of this sort were singular in the corpus.

Humor has strong ties to gender in the public eye. Women are said to have no sense of humor and to be bad comedians, which is nothing more than an androcentric stereotype. Perhaps it is also the fact that humor often deals with sexual themes and involves profanity that makes the stereotype so prevalent since it is more critical for a woman than a man to be modest and well-mannered (see section 1.1). Joking around is said to be one of the masculine features that boys learn at a young age [Tolson 1977, 32]. Mullany 2003's business meetings recordings show how women do not participate in the male banter and do not initiate humorous situations, rather doing the `interactional shitwork' [Fishman 1980] (see section 1.3) by supportively laughing at the jokes. It can be theorized than while there is no connection between having a sense of humor or being good at telling jokes and gender, men can be more encouraged to do so by one, the stereotype that they are indeed superior at it, and two, by knowing that their banter will be continued by other men as a bonding technique and/or supported by women who want to show support.

2.2.4.4 Self-directed function

A self-directed question “asks for information which only the speaker can supply, but orients the hearer toward the speaker in the process” [Freed 1994, 632]. Example (14) is cited from Freed's study:

(14) A: Well, think about it. I, what do I think? I think it's equal.

B: You do?

This resembles heavily conversational focus questions. In both cases, the information about to be conveyed is only available to the speaker. As stated in section 2.2.3.1, it is not evident why conversational focus questions should belong that high (relatively to the expressive style group) on the information continuum. Perhaps, these two functions could have been considered one. Unfortunately, lack of elaboration on Freed's part obscures her point which might have been legitimate. Both question types orient the hearer toward the speaker and pertain to information that only the speaker can provide. In relation to gender, not much can be proposed as far as this function goes.

2.2.4.5 Reported speech

Interrogative utterances that occur in recalling a separate (past or hypothetical) conversation, that is, are a quotation, are classified as reported speech questions.

Reported speech must not be equated to gossip; rather, gossip, as well as `troubles talk', `complaining', and `bitching' [Sotirin 2000] can all incorporate reported speech. What can be asserted is that recalling conversations (which was the majority of cases as opposed to hypothetical conversations) is characteristic of rapport-talk. According to [Tannen 1991], women prefer rapport-talk, whereas men do report-talk. Importantly, the opposition is not private versus public speaking (this deals with the communicative situation), rather personal versus impersonal [Tannen 1991, 261]. In other words, Tannen suggests that men and women are interested in different types of information and create connections based on different topics. This, of course, is a generalization, but one that can be taken as a basis for hypothesizing that a larger proportion of women's than men's questions would consist of retellings of past conversations as they pertain to private information.

3. Results

Table 1 depicts the structure of the resulting dataset.

See Appendices A and B for men and women's overall question distribution between functional types and groups.

Pearson's Chi-squared test has yielded a p-value of 2.818e-13 with a significance level of 0.05, meaning that the observed results were not due to chance and are statistically significant.

Fig. 2 represents a table of Pearson residuals, that is, visualizing the difference between expected and observed values. This allows to determine which discrepancies stand out the most and are deserving of more detailed analysis.

The most prominent difference in frequency is found with questions in reported speech. Women overall produced 191 such questions (7,6% of all their questions, see Appendix A), while men produced 94 (just 3,8%, see Appendix B). Closely follows the self-directed function with 38 and 77 questions or 3,1% and 1,5% for men and women respectively. Elaboration questions were more popular with men, who used 214 of them, while women did 152 (8,6% and 6,1% in turn). Some difference was found with public questions, men asking 236 and women asking 182 (9,4% and 7,3% respectively).

Since social and public questions were discussed in relation to each other and communicative styles that they are suggested to be characteristic of, a Chi-squared test was conducted. P-value= 0.006821, which means that the distribution of public and social questions between male and female speakers was statistically significant.

Less prominent but noticeable differences concern deictic and phatic questions, with men both times using them somewhat more frequently.

Fig. 3 depicts distribution of question types according to the communicative situation: same-sex or mixed-sex conversation.

The most difference between expected and observed values is seen with reported speech questions in mixed-sex (utterances addressed to women by men) and same-sex female communication. Namely, this question type occurs unexpectedly (according to predicted values) frequently in female to female conversations (9,1% of the times, see Appendix C) and less frequently than expected in male to female utterances (3,5%. See Appendix F). A Chi-square test for these values across communicative situations yielded a p-value of 0.03802, thus, proving a statistical significance of the results.

Self-directed function is observed to stray from the expected pattern by occurring more in female to female communication and somewhat less likely occurs in men's questions addressed to women. However, a Chi-square test for this parameter has shown no statistically significant correlation (p-value = 0.4737).

Elaboration questions are somewhat more likely to be addressed to women by men and occur less than expected occur in female same-sex talk. However, no statistical significance was found (p-value = 0.14).

Public information questions were used less than expected in female talk and just slightly more in male talk. Chi-square test has yielded a p-value of 0.05672, which is above the significance level of 0.05, showing no statistical significance.

Phatic information questions were used more in male talk than expected and slightly less popular than expected in female same-sex talk. However, p-value = 0.67.

Social invitation questions were used more than expected by women addressing men. Moreover, they were used less than expected in same-sex talk, both male and female. With p-value of 0.1406, no statistical significance is found.

Deictic questions were observed more frequently than expected in men's utterances addressed to women and less in same-sex conversations. P-value = 0.0004078, proving a statistical significance.

Social invitation questions, that is, indirect speech acts in form of questions, were more frequently than expected used by women in addressing men and less than expected in female same-sex talk. Chi-square test revealed a p-value of 0.0001815, meaning there is a statistical significance in the distribution.

Additionally, consider Table 2 showing the distribution of tag questions in the dataset.

Table 2 - Tag question distribution between genders

P-value calculated via the Chi-square test is < 0.00001, therefore, the proportions are statistically significant. Most notably, almost 3/4 of men's tag questions were directed at women, while women's tag questions are more evenly distributed according to the addressee's gender.

4. Discussion

4.1 Questions in reported speech

The most salient and curious finding was that both overall and in relation to the type of communicative situation, men and women's number of questions in reported speech varied. Women in total used more than twice as many such interrogatives. When it comes to the addressee's gender, female same-sex talk is the most saturated with this question type, comprising 9,1% of their interrogatives.

Before going into the discussion, it must be acknowledged that these data do not provide an insight on women's use of reported speech, only what role reported speech plays in accounting for interrogatives. From these results (see Appendices A through F) we can infer that a greater proportion of women's questions are pulled from outside (or, in some cases, imagined) conversations, which can be due to women's overall more frequent use of retellings. What is more, this question type is most used in female same-sex talk (9,1%, Appendix C) and least by men in mixed-sex talk (3,5%, Appendix F). Second highest use is by females in mixed-sex conversations (5,1%, Appendix E), followed by male same-sex talk (4,3%, Appendix D). It can be proposed that conversation retellings are most common for women that might prefer private talk or rapport-talk to public or report-talk (see section 2.2.4.5) since reported speech is cited primarily from personal experiences (rather than, say, a TV-show) . We could also theorize that women still stick to rapport-talk when talking to men, justifying this communicative situation's second highest percentage of questions in reported speech. Male same-sex communication is observed to contain fewer such interrogatives than female same-sex talk because men are speculated to adhere to rapport-talk, preferring discussing public matters instead of personal ones. However, at the same time, men use even less questions in reported speech when addressing women, which could be explained by the fact that men might be more willing to share personal matters with their sex that their friends happen to belong to as opposed to women that may not be included in their inner circle of friends. This hypothesis would have to be tested by closely examining the type of relationship that connects the speakers, as the nature of data collected for Spoken BNC2014 could, as is evident with gender distribution (see Table 1), be asymmetrical as it pertains to this factor.

Another issue that must be brought up here is that of gender salience. Hoggs' 1985 study of gendered speech in same-sex dyadic debates and mixed-sex groups employed a questionnaire asking outside listeners to evaluate to which degree speakers adhered to 10 stereotypically gendered characteristics. The results, as predicted, have revealed that women spoke in a more `feminine' way in same-sex dyads and more `masculine' in the groups. In men, this effect was less pronounced, but they, nonetheless, were said to have spoken more masculine in the groups and less so in dyads.

Hoggs bases his hypothesis on the `speech accommodation theory' [e.g., Thakerar et al. 1982] that proposes that “social approval is satisfied by convergence in language or speech style” [Hoggs 1985, 101], while “where speech style is a salient dimension of intergroup comparison” [ibid.], its divergence serves to highlight relative superiority. Hoggs claims that males “do not need to negotiate their superior status by means of extreme behavioural differentiation from females - it is already taken for granted, legitimate, and institutionalised” [ibid., 102].

While we operate different categories than Hoggs' study and our data are naturally occurring conversations instead of discussions, this model could be carefully applied to our findings. If we assume that women do indeed employ more retellings in their speech, they may be more encouraged to do so in same-sex talk that is focused on power symmetry and where such `female' linguistic features are not indicative of `lower' female status. Men, in their turn, could be less willing to retell private conversations (perhaps, perceiving this activity as heavily gendered `gossip') in mixed-sex talk because they want to seem more masculine. This speculation, however, to some degree contradicts the theory that males in their talk are always contesting the hierarchical status quo. One could argue that men would want to seem more masculine with other men through minimal usage of what they perceive as `gossip' (although, as discussed in section 2.2.4.5, gossip is not a solely female prerogative).

To sum up, frequent use of questions in reported speech in female same-sex conversation can be explained by women preferring private talk over public talk, while men stick to the latter, explaining their significantly lower numbers of such interrogatives. The distribution of these questions is also in accord with speech accommodation theory, but there are issues with possible explanations using this framework.

4.2 Deictic information questions

Deictic information overall accounted for 3,2% of men's questions and for 2,5% of women's. These percentages may not seem significant, but statistical analysis has shown that the distribution according to the communicative situation was not due to chance. In female same-sex talk, 2,1% (see Appendix C) of questions were of this type, in male same-sex talk 1,7% (see Appendix D). In mixed-sex conversation, 3,3% (see Appendix E) of questions were asked by women and 3,8% by men (see Appendix F).

As discussed before, speculation about who would be more interested or inclined to request deictic information would have to be based on at least some previous empirical evidence. Perhaps in this case, the results have more to do with the type of activities the speakers were engaging during the recording process: cooking, driving, playing videogames. Moreover, the pool of deictic questions may not have been sufficient to begin with, despite the chi-square test results. To conclude, an adequate theorical base in the cognitive field is needed in order to explain these findings as well as a bigger sample of this specific question type.

4.3 Social invitation questions

While men and women overall used a practically identical number of social invitation questions (76 and 73 respectively), their distribution according to the addressee's gender was uneven. The percentage was the same in same-sex talk (2,1% for both male and female, see Appendices A and B), but lower than in mixed-sex talk, where 3,3% (see Appendix E) of women's questions were social invitations and 3,5% (see Appendix F) were produced by men. Even though the chi-square test proves a statistical significance, the pool is once again less sufficient than would be desirable (see Appendices A through F). We can nevertheless infer that framing indirect speech acts as interrogatives is more common in mixed-sex conversations. Perhaps it is due to the perceived imbalance of power that can be observed in mixed-sex relationships that interlocutors aim to soften their commands and ask for permission (among other things, see section 2.2.1.2) in order to negotiate the asymmetry. However, as proposed by Coates (see section 1.3) or Tannen (see section 2.2.1.1), men tend to want to be in the position of dominance, so it would not be expected of them to be more polite (tying politeness with indirectness loosely) with women. It could be then speculated that they are comfortable with the distribution of power (slanted in their, men's, favor) and thus do not feel the need to establish dominance by implementing direct speech acts.

4.4 Tag questions

As discussed in section 2.2.3.3, tag questions have multiple context-sensitive purposes and may even display many functions at once. This research did not inspect tags in a detailed manner, but since, as a result of the study, quantitative data was already available, it was decided to consider the distribution that looked promising considering the relatively large sample.

Tag questions have become one of the most well-known and debated features that supposedly is characteristic of women's `weak' speech [Lakoff 1973]. Interestingly, men in our data were observed using these questions more frequently than women, 19,2% of all interrogatives were tags in men's speech compared to 16,2% for women (see Table 2). Their distribution according to the addressee's gender was proven to be of statistical significance. With both male and female speakers, most of their tags were addressed toward women, but female speaker's tags were more evenly distributed (57% were addressed to women), while outstanding 72,6% of men's tags were addressed to women.

Thus, this study proves once again that Lakoff's assumption regarding tag questions was not correct. In order to gain insight into the details of men and women's use of this syntactic feature, one must employ one of the proposed classifications regarding the tags' functions.?

Conclusion

This research has discovered that men and women differ the most in their use of questions embedded in reported speech. Women overall produced almost twice as many such interrogatives than men. The highest number was observed in female same-sex talk, followed by situations in which women addressed men, male same-sex talk, and situations where men addressed women. This pattern could be due to, as speculated, women preferring rapport-talk, and favoring this strategy still while talking to men. Men, on the other hand, might use a higher proportion of reported speech in same-sex talk because of the type of relationships they have with other men; that is, same-sex friendships are more frequent, and men feel more comfortable sharing past conversations with closer (male) friends. Further research thus might take into account the nature of speakers' relationship.

Differences in the frequency of use of deictic information questions, social invitations, and tag questions were also found. As for deictic questions, those were asked least in male same-sex talk, and most were employed by men addressing women. The reason for this distribution remains unclear and theoretical base in the cognitive field as well as further research using a bigger sample are required. Social invitations or indirect speech acts in interrogative form were used at the same rate by both genders but differed in their distribution according to the communicative situation. This question type was used most in mixed-sex conversation. A definite explanation cannot be proposed, so more research is needed.

Tag questions revealed to be used more by men, especially when addressing women. But since tags have long been divided by the function they take on in the conversation, this only gives an overall quantitative account of tag question use that, nevertheless, proves once again that tags do not characterize female speech.

Finally, it is important to notice that one must be careful generalizing about gender by drawing broad conclusions from one study. Even if trends are discovered, it is never the case that all members of said sex adhere to it. More research sensitive to the context, other parts of the speaker's identity like class or race, their goals and motivations inside the given conversation are key to building a clearer picture of the ways one chooses to perform gender linguistically.?

References

1.Bourdieu, P., and L. Boltanski. Le fйtichisme de la langue [Text] // Actes de la recherche en sciences sociales. - 1975. - Vol. 4. - P. 2-32.

2.Brown, P., and S. Levinson. Universals in Language Usage: Politeness Phenomena [Text] / E. Goody // Questions and Politeness: Strategies in Social Interaction. - Cambridge: Cambridge University Press, 1978. - P. 56-310.

3.Cameron, D. Gender and Language Ideologies [Text] / S. Ehrlich, M. Meyerhoff and J. Holmes. // The Handbook of Language, Gender and Sexuality. - 2014 (2003). - P. 281-296.

4.Carter, A. J., A. Croft, D. Lukas, G. M Sandstrom. Women's Visibility in Academic Seminars: Women Ask Fewer Questions Than Men [Text] // PLoS ONE. - Vol. 13(9).

5.Cheshire, J., and P. Trudgill. The Sociolinguistics Reader. Volume 2: Gender and Discourse [Text] / J. Cheshire and P. Trudgill. - 1998. - London: Arnold. - 416 p.

6.Coates, J. Women, Men and Language. A Sociolinguistic Account of Gender Differences in Language [Text] / J. Coates. - Harlow, England: Pearson Longman, 2013 (2004). - 246 p.

7.Eakins, B. W., and G. R Eakins. Sex Differences in Communication [Text] / B. W. Eakins and G. R. Eakins. - Boston: Houghton Mifflin, 1978. - 217 p.

8.Eckert, P. Variation and Social Practice: The Linguistic Construction of Social Meaning in Belten High [Text] / P. Eckert. - Wiley-Blackwell, 1999. - 240 p.

9.Eckert, P., and S. McConnell-Ginet. Think Practically and Look Locally: Language and Gender as Community-Based Practice [Text] // Annual Review of Anthropology. - 1992. - Vol. 21 (1). - P. 461-488.

10.Edwards, J. R. Social class differences and the identification of sex in children's speech [Text] // Journal of Child Language. - Vol. 6 (1). - 1979. - P. 121-127.

11.Fishman, P. Conversational Insecurity [Text] / H. Giles, W.P. Robinson, and P. M. Smith // Language: Social Psychological Perspectives. - Oxford: Pergamon Press. - P. 127-132.

12.Freed, A. F. The Form and Function of Questions in Informal Dyadic Conversation [Text] // Journal of Pragmatics. - 1994. - Vol. 21. - P. 621-644.

13.Freed, A. F., and S. Ehrlich. “Why Do You Ask?”: The Function of Questions in Institutional Discourse [Text] / A.F. Freed and S. Ehrlich. - Oxford: Oxford University Press, 2010. - 371 p.

14.Hayano, K. Question Design in Conversation [Text] / J. Sidnell and T. Stivers // The Handbook of Conversation Analysis. - Blackwell Publishing Ltd., 2013 - P. 395-414.

Appendices

dyadic gender interrogative

Appendix A - Female speakers' question distribution

Appendix B - Male speakers' question distribution

Appendix C - Female speaker's question distribution. Female addressee

Appendix D - Male speaker's question distribution. Male addressee

Appendix E - Female speakers' question distribution. Male addressee

Размещено на Allbest.ru